State of Innovation Summit

I'm at the Seed - Council on Competitiveness State of Innovation Summit. I was thinking about live blogging, but find that doing so makes it hard for me to think about what people are actually saying. There's a webcast if you're interested.

As far as conferences go, it's a good one. Rock stars on the stage (E.O. Wilson is a hero of mine) and interesting conversations about innovation.

But I'm frustrated, as I often am at "innovation" conferences. What follows is a bit of a rant directed less at this event, which as I said is a good one, but at the conversation I hear all the time about scientific innovation. There are three problems.

Problem 1: there's almost no conversation about the essential theories of emerging innovation - open, user-driven, distributed. This is about the new forms of innovation that the network enables, and should be on every agenda of every meeting that claims to talk about innovation. If we simply do things the old way, but bigger, we fail. Disruptive innovation models ought to be part of the conversation and they too often aren't.

Problem 2: there's no conversation about technical infrastructure for innovation. Here's what I mean by that: the internet is infrastructure for innovation in culture and commerce. It underpins an enormous amount of economic value, and from it emerges disruption that we could never have predicted, like the Web. And the web in turn begat Google, Amazon, Facebook, blogging, you name it. Both of these systems work this way because they are public systems. Yet we don't talk about an open public technical infrastructure for science. We build individual bits of it, but our vision is investing in unconnected nodes, not networks.

On top of this, there is the assumption that because the web works for culture, it works for science. But the Web is a system built for documents - it's infrastructure for documents. Science innovation depends on data. This conference had a great panel on data, with Ben Fry, who's a data visualization wizard. Yet no conversation that the infrastructure we have for the Web completely fails at data. Infrastructure for making the web function on data is woeful - format standards, annotation, and so on are always underfunded and first to cut in crisis.

Infrastructure for data integration, data federation, and so forth should be encoded directly into the open standards of the web and internet. Full stop. And we should talk about this problem more often. Otherwise people look at their iPhones, check for a latte, and assume this level of functionality scales from coffee to the bench. It doesn't.

Problem 3: there's no conversation about the way that our legal and policy regimes affect emerging modes of innovation. Data use is dependent on legal access to data. There's a range of data regimes across the world that make legal access to data conditional on rights being granted. Copyright licenses prevent innovative scientists from using software to index the literature and integrate it into the database world. Default settings on government policy create strong incentives for patenting smaller and smaller inventions by universities. Tenure and review systems encourage secrecy and withholding.

Taken together, these three problems represent the core "immune system" of science to disruptive change. That's not a terrible thing. Science should resist some disruptive changes. But right now, the disruptive change being resisted is the network. It's a terrible irony that at the moment we have the technical ability to send any content anywhere at almost no cost of distribution, we haven't got the technical and legal infrastructure to realize the potential of that ability for science. It's an even more terrible irony that the innovation resulting from that ability in culture is being constricted by the very policies and regimes we claim to promote innovation.